TY - JOUR
TI - Deep learning technology for face recognition
AU - Yuan Bin
JN - Thermal Science
PY - 2025
VL - 29
IS - 3
SP - 2007
EP - 2014
PT - Article
AB - In China, the rapid development of public transportation network construction  has been accompanied by a high incidence of traffic accidents caused by  sleep-deprived driving. The monitoring of drivers' sleep-deprived driving  and the sending out of early warnings has been identified as a field of  research with both important theoretical and practical value. This article  proposes a fatigue detection algorithm based on facial recognition  information fusion. The algorithm extracts facial feature information and  head features from the driver's face and fuses them into facial recognition  information parameter feature vectors. A set of feature nodes is thus  constructed, and an information fusion feature map is constructed according  to the interaction information between nodes. The fatigue status is  detected using the label propagation strategy, which is more effective than  other benchmark detection algorithms. The fatigue detection algorithm has  practical value.